Beta. Content is under active construction and has not been peer-reviewed. Report errors on
GitHub
.
Disclaimer
Theorem
Path
Curriculum
Paths
Demos
Diagnostic
Search
Quiz Hub
/
Gaussian Mixture Models and EM
Gaussian Mixture Models and EM
3 questions
Difficulty 3-6
View topic
Foundation
0 / 3
1 foundation
2 intermediate
Adapts to your performance
1 / 3
foundation (3/10)
state theorem
A Gaussian Mixture Model (GMM) represents data as a weighted sum of
K
Gaussians. The EM algorithm alternates two steps. What are they?
Hide and think first
A.
Gradient step on the log-likelihood, then a projection step to keep covariances positive definite.
B.
Merging clusters that are close, then splitting clusters that have high variance.
C.
Random sampling of points, then counting points per cluster to estimate mixing weights.
D.
E-step: compute posterior probabilities of cluster assignments. M-step: re-estimate Gaussian parameters (means, covariances, mixing weights) using those posteriors as soft labels.
Submit Answer